This is an admittedly unusual thing to do, but I typed
'cat /dev/zero > /dev/sd1c' and got this:

     Dec  7 02:29:39 bullbox /netbsd: sd1(ncr0:1:0): volume overflow, info = 586764 (decimal), data = 00 00 00 00 21 00 00 00 00 00
     Dec  7 02:29:39 bullbox /netbsd: sd1(ncr0:1:0): illegal request, data = 00 00 00 00 21 00 00 c0 00 02
     Dec  7 02:30:10 bullbox last message repeated 299 times

I deleted numerous "last message..." lines, since I started this
and then went to bed for the night.  In the morning, I had to press
the reset button to get my machine back after seeing about a hundred
lines saying things like last message repeated 6173 times (they
were all around 6100 repeats).  I wanted to write to every sector
of the disk, to get the defect list built up, and I figured that
the cat was the easiest way to do it.  (I was getting soft errors
that were corrected, and I wanted the drive to reallocate the bad
blocks.)

The reverse procedure 'cat /dev/sd1c > /dev/null' produced:

     Dec  7 08:47:04 bullbox /netbsd: ncr.c: timeout counter = 999943, len = 2048 count=4096 (count-len 2048).
     Dec  7 08:47:04 bullbox /netbsd: ncr_debug = 2,  1=out, 2=insd1(ncr0:1:0): illegal request, data = 00 00 00 00 21 00 00 c0 00 04
     Dec  7 08:47:05 bullbox /netbsd: ncr.c: timeout counter = 999943, len = 2048 count=4096 (count-len 2048).
     Dec  7 08:47:05 bullbox /netbsd: ncr_debug = 2,  1=out, 2=insd1(ncr0:1:0): illegal request, data = 00 00 00 00 21 00 00 c0 00 04
     Dec  7 08:47:05 bullbox /netbsd: sd1(ncr0:1:0): illegal request, data = 00 00 00 00 21 00 00 c0 00 02

With no real significant problems.  (I didn't even have to kill
the process, as it said "cat: /dev/sd1c: Invalid argument" and
died.)  I'm not real concerned about this, but thought someone
might want to address them if they are in fact deemed to be real
bugs.  I was operating under the idea of "Kids, don't try this at
home" and was expecting to see unusual activity, due to the hardware,
if not the operating system on top of it.

Jon Buller